NEW YORK – Uptake of Accelerate Diagnostics' Pheno system by clinical microbiology labs has been sluggish, but initiatives soon to be led by a newly appointed chief operating officer combined with new clinical outcomes data could lead to a tipping point for the Tucson, Arizona-based firm.
While the firm views the new outcomes data as a positive, investment analysts gave mixed reviews. Specifically, Brian Weinstein at William Blair enthusiastically maintained an Outperform rating for Accelerate's stock, calling the new data "clearly positive," and saying it will "help set the table" for "a breakthrough 2020" for the company. But Tycho Peterson at JP Morgan saw the glass as half empty, calling the results "a clear disappointment," and downgrading the company's stock from Neutral to Underweight.
Accelerate Diagnostics believes it has a game-changing technology to identify infectious disease pathogens and determine phenotypic resistance characteristics much faster than traditional methods, the firm's CEO Larry Mehren asserted in an interview this week.
The firm's Pheno system and bloodstream infection assay were cleared by the US Food and Drug Administration in February of 2017. Initial sales of the instruments and tests were slow, but the firm switched to a reagent rental model in mid-2018 in order to place systems free of cost, selling the tests for approximately $200 each. Accelerate's placement of 78 instruments in 2017 grew to 202 in 2018, and, combining the 2019 first and second quarter results, it has placed 114 systems in the first half of this year.
With the new COO, Jack Phillips, and new data, "I believe that we are at the tipping point at Accelerate," Mehren said. Phillips previously served as CEO and president of Roche North America, and Mehren noted that in this position he took Roche from a ranking of number three or four among diagnostics makers in the US, "behind Abbott and Siemens," to the front of the pack.
"To get a person like that is rare," Mehren said, adding that he expects Phillips to "absolutely industrialize the sales and installation process" so that Accelerate can "drive a repeatable sales process globally."
Establishing a strong sales process has been a particular bottleneck for Accelerate, perhaps due in part to the complexity of its test and system.
In a recent conference call to discuss quarterly earnings, Mehren noted that the installation process has taken customers many months. "There is a lot of opportunity for us to shorten that timeframe beyond where we are today, which is just too long," he said this week, noting that it currently takes customers more than six months to get Pheno up and running at their labs.
"We have, in my opinion, one of the most impactful diagnostic devices ever released into the market, we just need to match that innovation with a very innovative and repeatable sales and installation process, and that's why I brought Jack in," Mehren added.
The firm is now planning on expanding commercial efforts in the US, in particular, as well making its first hires in China, specifically market development people who will ultimately be working with distributors and supporting their efforts in that country, Mehren said. The firm is also seeing "significant uptake" of its system in the Middle East and in Eastern Europe, growing uptake in Greece and Germany, and continued strength in Italy and Spain, he said.
Accelerate will submit an application for a New Technology Add-On Payment to the Centers for Medicare and Medicaid Services next month, Mehren said. T2 Biosystems became the first diagnostics maker to get an NTAP earlier this month, and the payment is expected to incentivize uptake of T2's system by reducing the cost of its test for Medicare patients by approximately 65 percent.
Interpretations differ on new outcomes data
Outcome studies for the Pheno system have been highly anticipated by labs and investors. Mehren earlier this year described the firm's future strategies of emphasizing positive outcomes data in its sales pitches to hospitals.
Romney Humphries, CSO at Accelerate, explained that it has already been established that the Pheno test performs a rapid identification and determination of antibiotic susceptibilities from a positive blood culture, "providing data about 40 hours quicker than you could do by traditional methods."
The question for the outcomes studies has been, "How does that impact patient management?," she said.
Three new outcomes studies of the Pheno will be presented at the upcoming IDWeek conference, and abstracts were posted online last week.
In a nutshell, the data showed the test can help clinicians put patients with sepsis on the optimal antibiotic sooner, with a particular advantage over gold-standard culture-based methods for a subset of the pathogens in the panel. But, in one study that looked at so-called secondary outcomes, the benefits did not seem to extend to patient mortality or lead to a reduction in how long the patient had to stay in the hospital.
Specifically, a multi-center randomized controlled trial by researchers at the Mayo Clinic and the University of California, Los Angeles, showed an approximate six-hour decrease overall in the time it took to appropriately adjust empiric antibiotic therapy, and more than 24 hours of improvement specifically for Gram negative pathogens. Differences in mortality and length of stay did not achieve statistical significance, but Accelerate noted that the study was not designed to include a sufficient number of patients to statistically evaluate these endpoints.
The study, called Rapid Identification and Antimicrobial Susceptibility Testing for Gram-Negative Bacteremia, or RAPIDS-GN, focused on Gram negative bacteria.
Researchers at the University of Iowa, meanwhile, showed that antibiotic therapy was optimized within about 15 hours, and 80 percent of patients had an antibiotic change as a result of the Pheno test results. In about one-third of these cases, the change was needed because the empiric antibiotic didn't cover the pathogen the patient was infected with.
Researchers from the University of Arkansas Medical System found an approximate one-day improvement in time to optimal therapy with the Pheno system. In that study, the researchers compared results to historic cohorts and found between a half-day to two-and-a-half days shorter length of stay for patients tested with Pheno.
The Mayo/UCLA study focused on Gram negative pathogens in part because these had been a sticking point in previous evaluations of other rapid tests for bloodstream infections.
In a 2015 study, the Mayo Clinic researchers showed that physicians only altered their prescribing after being given genetic identification and resistance results from the BioFire BCID test when there was also an antibiotic stewardship intervention, and the testing had no impact on mortality, length of stay, or cost. A 2018 study also showed that the rapid test had little effect on prescribing behavior for Gram negative bacteria in particular.
"The hypothesis was that, for Gram negatives, they are so complex and there's a lot of antibiotic resistance, so people really needed the phenotypic susceptibility results in order to take action," Humphries said.
The current study didn't question whether stewardship was required, and didn't use any enhanced stewardship approaches, Humphries said. "In fact, the study really looked at the pragmatic real-world situation where, if the results were available in the middle of the night, they would act on them the next morning, like you would in the real world," she said.
Generally speaking, there seems to be a push from customers and payors for companies to demonstrate that their tests have an impact on patient management, health, and overall costs. These studies have to be done by clinical labs in the trenches and preferably at multiple centers to get large enough sample sizes.
"As lab directors, we've always really focused in on the analytical performance of a test, asking the question, 'Does it work according to how it was cleared or what the claims are,'" Humphries said. "That of course is important — you don't want to be running a test that is going to give results that are erroneous — but the outcomes data really takes it to the next step," she said, adding, "You're putting out a new result. Do people take action on that result? Is it helpful to them in the management of their patients? And a step further than that is saying, 'Does it make an impact to that patient's outcome?'"
These primary and secondary outcomes, however, might not be so clear cut in a population of complicated patients, like those with sepsis.
"I think one thing that we're learning is that the more steps away from the individual test you take, the more noise there is," Humphries said, and with more noise in the results, it will take larger sample sizes to see statistically significant differences.
"In some cases, while maybe we want to show mortality benefits or length-of-stay benefits, some of these patients are really complicated, and their infection is one part of what brings them to the hospital," she said, noting that some of these patients become septic after an organ transplant or a car accident, and their length of stay or morality might not be entirely dependent on getting their bloodstream infection under control.
William Blair's Weinstein took a likewise nuanced stance in an investment note with respect to the secondary outcomes.
He emphasized that the Mayo/UCLA study was only powered to show the difference in the time to antibiotic modification, but it "was not powered to drive a statistical conclusion on patient outcomes such as length of stay, morbidity, mortality, stewardship, and costs, though we believe trend data on these measures were also positive," he wrote.
Indeed, Mehren also said that the RAPIDS study, in particular, demonstrates that "even at the world's best institutions" that have low resistance rates and excellent stewardship programs, the Pheno can impact patient management "in a very positive, important way" by getting patients on the right antibiotic much sooner.
"If we can do that at Mayo Clinic … we can do that anywhere," he said.
But, JP Morgan's Peterson said in a note that the Mayo/UCLA study "failed to meet key secondary endpoints, which are more important than the primary endpoint, in our view."
In an investment note released today, he emphasized that secondary endpoints are more important than the primary endpoint specifically "in demonstrating the value proposition of the Pheno system."
He further said that, "Although management attributed this to the fact that Mayo has among the lowest gram-negative resistance rates in the US and already has long-standing antibiotic stewardship and sepsis management programs in place, hence making it more difficult to show incremental improvement over a high baseline, this was still a clear disappointment [versus] prior expectations and certainly does not help as the company pushes to meet its ambitious 2H placement goals."